Too often, tests of economic rationality of agricultural producers are based on the assumption of profit maximization and certainty of outcomes of production decisions (Hopper; and Yotopoulos). However, multiple goals and uncertainties may be relevant to the decisionmaker. Consequently, single goal models under certainty are not always a realistic approach to the decision process and may not provide a farmer with an acceptable solution. Progress in using multiple goals and uncertainties in decision models has been slowed by the difficulty of incorporating multiple goals and uncertainties into analytical models. However, with the development of decision theory, procedures became available that permit explicit incorporation of uncertainties and multiple goals. Decision theory describes how a rational decisonmaker ought to behave given his beliefs and preferences. Whether or not the model explains the behaviour of peasant farmers can only be answered by empirical tests. The objectives of this study are: (1) to compare alternative theories of choice (single attribute utility maximization, multiattribute utility maximization, and expected profit maximization in terms of their abilities to explain and predict actual resource allocation decisions of producers, and (2) to explore implications for policy decisionmakers of the impact of and uncertainty on farmers' choices. The present research was conducted on a sample of rice farmers in Sri Lanka.